CPV301

[QUES] The input of computer vision is an image. So what is the output of computer vision?

A. The output is the interpretation of an image.

B. The output is a processed image.

C. The output is the image that has been recovered with the parts that have been noisy-

D. All of the others

A. The output is the interpretation of an image.

[QUES] What system does the digital camera use to perform the digitization of images?

A. Number system.

B. Logic system

C. Sensor system

D. Optical system

D. Optical system

[QUES] In K-means for Segmentation, what is the meaning of K value?

A. It represents the number of clusters.

B. It represents the number of images.

C. It represents the step of clusters.

D. It represents the color groups.

=> A. It represents the number of clusters.

[QUES] Lambert's cosine law is used to calculate the intensity of diffuse reflection. What parameter is important in Lambert's cosine law?

A. The intensity of the light emitted by the light source.

B. The angle between the incident light direction and the surface.

C. The type of matter that makes up the surface of matter, affecting absorption or reflection.

D. The calculated spatial coordinates may change as you convert between spatial domains.

=> B. The angle between the incident light direction and the surface.

[QUES] What is the main approach of feature-based alignment?

A. Align the 4 corners of two photos

B. Finding a way to recover an image's angle through spatial transformations

C. Overlap the pictures together and find similarities between them

D. Find a few matching features in both images then compute alignment

=> D. Find a few matching features in both images then compute alignment

[QUES] The..... is a measure of change in a function, and an image can be considered to be an array of samples of some continuous function of image intensity- Choose the best answer.

A. Gradient

B. Edge

C. Grayscale

D. Contour

A. Gradient

[QUES] Choose the best description of the Prewitt operator for edge detection

A. It detects the overlap edges of an image.

B. It is one of the best ways to detect the color intensity and magnitude of an image

C. It is a good performance on detecting vertical and horizontal edges

D. It is never a good choice to detect the orientation of an image

C. It is a good performance on detecting vertical and horizontal edges

[QUES] Choose the best description about bundle adjustment in image stitching

A. Bundle adjustment is still needed to obtain a globally coherent mosaic.

B. The goal of Bundle adjustment is then to find a globally consistent set of alignment parameters that minimize the mis-registration between all pairs of images.

C. Bundle adjustment is the process of simultaneously adjusting pose parameters for a large collection of overlap- ping images.

D. Bundle adjustment refers to the problem of finding the image registration parameters that best comply with the constraints introduced by the image matching.

C. Bundle adjustment is the process of simultaneously adjusting pose parameters for a large collection of overlap- ping images.

[QUES] The aim is to identify the facial landmarks that are key to distinguishing your face. What is the exact name of this step in face recognition? Choose the best correct answer

A. Face detection

B. Face locating

C. Features analysis

D. Face analysis

D. Face analysis

[QUES] Choose the best correct answer about parallax removal

A. Parallax is a key challenge that leads to inaccurate registration and ghosting effect of objects in the result. So it needs to remove.

B. It is used to optimize the camera poses. Once these have been estimated, the desired location of a 30 point xi can be estimated as the average of the back-projected 3D locations.

C. It is a displacement or difference in the apparent position of an object viewed along two different lines of sight, and is measured by the angle or semi-angle of inclination between those two lines.

D. All of the others

C. It is a displacement or difference in the apparent position of an object viewed along two different lines of sight, and is measured by the angle or semi-angle of inclinat ion between those two lines.

[QUES] We have source image f(x) and transformation formula x'= h(x). We can compute the values of the pixels in the new image g(x) by using MIP-mapping. Choose the best description about MlP-mapping

A. A MlP-map is a standard image pyramid where each level is pre-filtered with a high-quality filter rather than a poorer quality approximation.

B. A MlP-map are pre-calculated, optimized sequences of images, each of which is a progressively lower resolution representation of the previous.

C. All of the others

D. None of the others

A. A MlP-map is a standard image pyramid where each level is pre-filtered with a high-quality filter rather than a poorer quality approximation.

[QUES] ....is a technique of image processing used to identify points in a digital image with discontinuities ~ sharp changes in the image brightness. Choose the best answer

A. Peak detector

B. Peak descriptor

C. Edge detection.

D. Edge descriptor

=> C. Edge detection.

[QUES] Choose the best description of the Gaussian Mixture Model for segmentation

A. It assumes that there are a certain number of Gaussian distributions, and each of these distributions represents a cluster.

B. It tends to divide the data points belonging to a single distribution together.

C. It is used to identify different classes or clusters in the given data based on how similar the data is.

D. All of the others

A. It assumes that there are a certain number of Gaussian distributions, and each of these distributions represents a cluster.

[QUES] What are languages supported by Computer vision?

A. C/ C++

B. Java

C. Python

D. All of the others

D. All of the others

[QUES] It is a type of segmentation technique that can be defined as the use of energy forces and constraints for the segregation of the pixels of interest from the image. What is the exact name of this process?

A. Active contour

B. Contour detection

C. Multi-pass filter

D. Boundary detection

A. Active contour

[QUES] What happens in the thin lens model?

A. The object is inverted.

B. The image is discolored.

C. Objects placed before or after Len produce the same results-

D. The image will stay the same size when you use white light as the source (sun).

A. The object is inverted.

[QUES] The most obvious application is used in scanning large documents - beyond the size of the scanner. What is the best solution?

A. Use transformations to shrink documents before scanning

B. Scan part of the image then uses interpolation methods-

C. Use the technique to find important features in the document, then perfomt a scan and use extrapolation to find the missing parts.

D. Use multiple scanners, scan parts of the document, and then stitching them together.

D. Use multiple scanners, scan parts of the document, and then stitching them together.

[QUES] It is the task of taking an initial set of object detections, creating a unique ID for each of the initial detections, and then tracking each of the objects as they move around frames in a video, maintaining the ID assignment.

What is the exact name of this process? Choose the best correct answer

A. Object detection

B. Object locating

C. Object tracking

D. Object recognition

C. Object tracking

[QUES] A convolution-based technique is used to detect the line. Here are the four convolution masks to detect

horizontal, vertical, oblique +- amount lines in an image. What is the specific value of the oblique in degrees?

A. 15

B. 45

C. 90

D. 180

B. 45

[QUES] What do the components of a pixel include?

A. Intensity value and size of the image

B. Position address and intensity value

C. Position address and size of the image

D. All of the others

B. Position address and intensity value

[QUES] What is the use of the geometric transformations in real life?

A. Align multiple images that were clicked at different times.

B. Correct images for avoiding lens distortion.

C. Images according to camera orientation-

D. All of the others

D. All of the others

[QUES] Choose the best description of the edge-based segmentation

A. Edge-based segmentation algorithms work to detect objects in an image.

B. it bases on various continuities in grey level, colour, texture, brightness, saturation, contrast etc.

C. These operators aid in detecting the edge discontinuities and hence mark the edge boundaries.

D. it is a type of image segmentation, where we change the pixels of an image to make the image easier to

analyze.

A. Edge-based segmentation algorithms work to detect objects in an image.

[QUES] Why use the image feature for processing, rather than using the whole image

A. In the photo, we are only interested in the color, as well as the size of the photo.

B. Use features to be able to apply more advanced techniques later like deep learning.

C. It reduces the number of resources required to describe a large set of data

D. All of the others

D. All of the others

[QUES] Choose the best description of the active contours

A. It is the process of assigning a color to every pixel in an image.

B. it defines a separate boundary or curvature for the regions of the target object for segmentation.

C. Image segmentation is typically used to recognize the color range and their group in images-

D. It is described as a multi-model for the process of segmentation

B. it defines a separate boundary or curvature for the regions of the target object for segmentation.

[QUES] The input of computer vision is an image. 80 what is the output of computer vision?

A. The output is the interpretation of an image.

B. The output is a processed image.

C. The output is the image that has been recovered with the parts that have been noisy.

D. All of the others

A. The output is the interpretation of an image.

[QUES] ln image processing, what is the use of the digital camera?

A. The digital camera is used to convert digital images to images stored in the computer.

B. The digital camera is a camera that captures photographs in digital memory

C. The digital cameras use an arithmetic system, for the acquisition and representation of images.

D. The digital camera is a device that captures input and output images.

B. The digital camera is a camera that captures photographs in digital memory

[QUES] Choose the best answer about homogeneous coordinates

A. Homogeneous coordinate is a system of coordinates used in projective images.

B. It is used to change the color of the original image to a new dimension.

C. It is used to perform salt and pepper noise filters.

D. It is the way to approach the projective plane analytically

D. It is the way to approach the projective plane analytically.

[QUES] What is computer vision?

A. Computer Vision is a field of computer science that enables the computer to understand just like a human.

B. Computer Vision is the analysis and manipulation of a digitized image, especially in order to improve its quality.

C. Computer \fision is the study of computers and computational systems like electrical and computer engineers.

D. Computer Vision is a task defined as the technical analysis of an image by using complex algorithms.

A. Computer Vision is a field of computer science that enables the computer to understand just like a human.

[QUES] Choose the best description of the Gaussian Mixture Model for segmentation

A. it assumes that there are a certain number of Gaussian distributions, and each of these distributions

represents a cluster.

B. It tends to divide the data points belonging to a single distribution together.

C. It is used to identify different classes or clusters in the given data based on how similar the data is.

D. All of the others

A. it assumes that there are a certain number of Gaussian distributions, and each of these distributions

[QUES] In K-means tor Segmentation, what is the meaning of K value?

A. lt represents the number of clusters.

B. It represents the number of images.

C. It represents the step of clusters.

D. lt represents the color groups.

A. lt represents the number of clusters.

[QUES] Choose the best description of the active contours

A. It is the process of assigning a color to every pixel in an image.

B. It defines a separate boundary or curvature for the regions of the target object for segmentation.

C lmage segmentation is typically used to recognize the color range and their group in images.

D. It is described as a multi-model for the process of segmentation.

=> B. It defines a separate boundary or curvature for the regions of the target object for segmentation.

[QUES] It is a type of image segmentation, where we change the pixels of an image to make the image easier to analyze. We convert an image from color or grayscale into a binary image, i.e-, one that is simply black and

white. What is the exact name of this process?

A. Thresholding Segmentation.

B. Edge-Based Segmentation.

C. Region-Based Segmentation.

D. Watershed Segmentation.

A. Thresholding Segmentation.

[QUES] Choose the best description of the Hough transform

A. The Hough transform is a technique that can be used to isolate features of a particular shape within an image.

B. Hough transform can be employed in applications where a simple analytic description of a feature is not possible.

C. The main advantage of the Hough transform technique is that it is tolerant of gaps in feature boundary descriptions and is relatively unaffected by image noise.

D. All of the others

D. All of the others

[QUES] Choose the best description of the gaussian smoothing

A. Gaussian smoothing is the result of blurring an image by a Laplace function.

B. It enhance image structures at different scales.

C. Smooth blur resembling that of replace the image through a translucent screen

D. All of the others

B. It enhance image structures at different scales.

[QUES] It is a type of image segmentation, where we change the pixels of an image to make the image easier to analyze. We convert an image from color or grayscale into a binary image, i.e., one that is simply black and

white. What is the exact name of this process?

A. Thresholding Segmentation.

B. Edge-Based Segmentation.

C. Regionâ"Based Segmentation.

D. Watershed Segmentation.

A. Thresholding Segmentation.

[QUES] In image processing, why is the Fourier transform used?

A. The Fourier Transform is used ifwe want to access the geometric characteristics of a frequency domain image.

B. Because the image in the Fourier domain is decomposed into its sinusoidal components.

C. It is easy to examine or process certain spatial of the image, thus influencing the geometric structure in the spatial domain.

D. All of the others

D. All of the othe

[QUES] Why do we need image stitching?

A. Image stitching is a widely used technique for recovering original data from ripped data.

B. Image stitching is used in forensic and investigative science for the reconstruction of tom paper which is a big problem.

C. In image mapping, stitching of images is done to do the complete mapping of a particular place-

D. All of the others

=> D. All of the others

[QUES] Algorithms used for face detection based on what set features to recognize faces? Choose the best correct

answer

A. Eyes, eyebrows, nose, hair, jaw

B. Eyes, eyebrows, nose, neck, mouth

C. Eyes, eyebrows, nose, mouth, jaw

D. Eyes, eyebrows, jaw, neck, ear

C. Eyes, eyebrows, nose, mouth, jaw

QUES] What is a geometric primitive?

A. They work just like the blocks in a typical preschool building set

B. Geometric primitives are represented as points and lines.

C. Geometric primitives are the basic shapes we all know and recognize

D. All of the others

D. All of the others

[QUES] What is the basic principle of object detection?

A. Object detection is based on the features that make up the object.

B. Each object's surface has a different structure and shape. Hence object detection relies on color relevance.

C. Object detection is based on the semantics of the objects in the image.

D. All of the others

[QUES] What is the basic principle of object detection?

A. Object detection is based on the features that make up the object.

B. Each object's surface has a different structure and shape. Hence object detection relies on color relevance.

C. Object detection is based on the semantics of the objects in the image.

D. All of the others

[QUES] Choose the best description of the feature descriptors

A. Feature descriptor is an algorithm that takes an image and outputs feature vectors.

B. Feature descriptors are used to find the essential features from the given image

C. Feature descriptors are used to convert features to other representations through spatial transformations.

D. All of the others

A. Feature descriptor is an algorithm that takes an image and outputs feature vectors.

[QUES] Choose the best description of the Shearing transfonnations

A. Shearing an image means shifting the pixel values either horizontally or vertically.

B. It is sounding like translation, but the difference is that shifting is not done by the same amount.

C. Basically, this shifls some part of an image to a direction and other parts to some other direction.

D. All of the others

C. Basically, this shifls some part of an image to a direction and other parts to some other direction.

[QUES] After extracled features and their descriptors from two or more images then the next is to establish some

preliminary feature matches between these images. Which name is it called?

A. Feature reduce

B. Feature combination

C. Feature compression

D. Feature matching

D. Feature matching

[QUES] Choose the best answer about digital image

A. The cfigital image is a two-dimensional array of color

B. The digital image is a matrix of pixels.

C. The digital images are composed of a combination of colors.

D. The digital images are created from digital cameras through a mechanical system.

B. The digital image is a matrix of pixels.

[QUES] Choose the best description of the Hough transform

A. The Hough transform is a technique that can be used to isolate features of a particular shape within an image.

B. Hough transform can be employed in applications where a simple analytic description of a feature is not possible.

C. The main advantage of the Hough transform technique is that it is tolerant of gaps in feature boundary descriptions and is relativer unaffected by image noise.

D. All of the others

A. The Hough transform is a technique that can be used to isolate features of a particular shape within an image.

[QUES] How are thresholding methods implemented? Choose the best answer

A. lt replaces each pixel in an image with a black pixel if the image intensity is less than some fixed constant.

B. It replaces each pixel in an image with a white pixel if the image intensity is greater than that constant.

C. The first and second is correct

D. The first or second is correct

C. The first and second is correct

[QUES] In ........, the value of an output pixel is also computed as a weighted sum of neighboring pixels.

A. Filter

B. Coneiation

C. Kernel

D. Point

A. Filter

[QUES] Choose the best description of the watershed segmentation

A. A watershed is a transformation defined on a grayscale image.

B. The watershed transform decomposes an image completely and thus combine each pixel either to a region

C. The morphological operator watershed uses regional maximize as seeds for segmenting regions.

D. Watershed segmentation is a content-based technique that utilizes image morphology

A. A watershed is a transformation defined on a grayscale image.

QUES] Choose the best description of the K- Mean for segmentation

A. The k-means clustering algorithm attempts to split a given anonymous data set into 3 dynamic number of clusters.

B. It is used to identity different classes or clusters in the given data based on how similar the data is.

C. K Means is a classitying algorithm.

D. Data points in the neighbor group are more similar to other data points in that neighbor group than those in other groups.

B. It is used to identity different classes or clusters in the given data based on how similar the data is.

[QUES] Lambert's cosine law is used to calculate the intensity of diffuse reflection. What parameter is important in

Lambert's cosine law?

A. The intensity of the light emitted by the light source.

B. The angle between the incident light direction and the surface.

C. The type of matter that makes up the surface of matter, affecting absorption or reflection.

D. The calculated spatial coordinates may change as you convert between spatial domains.

B. The angle between the incident light direction and the surface.

[QUES] What is the main approach of feature-based alignment?

A. Align the 4 corners of two photos

B. Finding a way to recover an image's angle through spatial transformations

C. Overiap the pictures together and find similarities between them

D. Find a few matching features in both images then compute alignment

D. Find a few matching features in both images then compute alignment

[QUES] The is a measure of change in a function, and an image can be considered to be an array of samples of some continuous function of image intensity. Choose the best answer

A. Gradient

B. Edge

C. Grayscale

D. Contour

A. Gradient

[QUES] Choose the best description of the Prewitt operator for edge detection

A. lt detects the overlap edges of an image.

B. It is one of the best ways to detect the color intensity and magnitude of an image

C. It is a good performance on detecting vertical and horizontal edges

D. It is never a good choice to detect the orientation of an image

C. It is a good performance on detecting vertical and horizontal edges

[QUES] Choose the best description about bundle adjustment in image stitching

A. Bundle adjustment is still needed to obtain a globally coherent mosaic.

B. The goal of Bundle adjustment is then to find a globally consistent set of alignment parameters that minimize the mis-registration between all pairs of images.

C. Bundle adjustment is the process of sim ultaneously adjusting pose parameters for a large collection of overlap- ping images.

D. Bundle adjustment refers to the problem of finding the image registration parameters that best comply with the constraints introduced by the image matching.

C. Bundle adjustment is the process of sim ultaneously adjusting pose parameters for a large collection of overlap- ping images.

[QUES] To recognize panoramas. Brown and Lowe (2007) first find all overlaps using a featureâ"based method and then find connected components in the overlap graph to "recogni7e" individual panoramaa

A. some images

B. single Image

C. feature-based

D. paimise image

D. paimise image

[QUES] 112. Which of the following comes under the application of image blurring?

A. Image segmentation

B. Object motion

C. Object detection

D. Gross representation

D. Gross representation

[QUES] What is the name of the process that moves a filter mask over the image. followed by calculati ng the sum of products?

A. Linear spatial flltering

B. Correlatlon

C. Convolutr'on

D. Non-linear spatial filtering

C. Convolutr'on

[QUES] Which of the following operations Is used in homographic f iltering for converting the input image to discrete Fourier transformed function?

A. Logarithmic Function

B. Negative Function

C. filter Function

D. Exponential Function

A. Logarithmic Function

QUES] The DIFS and DFFS metrics first developed by Moghaddam and Pentland (1997) are used to produce_ Mahalanobis distance measurements.

A. 24

B. 28

C. 32

D. 20

B. 28

[QUES] DIFS stands for (Moghaddam and Pentland 1997).

A. dive in face space

B. distance in face space

C. deep in face space

D. dense in face space

B. distance in face space

[QUES] lAlgorithm:

-Define the criterion to be used for homogeneity

â"Split the image into equal size regions

â"Calculate homogeneity for each region

â"If the region is homogeneous, then merge it with neighbors

â"The process is repeated until all regions pass the homogeneity test

A. Watershed

B. Region Splitting

C. Split and merge Segmentation

D. Graph-based Segmentation

C. Split and merge Segmentation

[QUES] Finding faces and allowing users to tag them makes it easier to find photos of selected people at a later date or to automatically share them with friends. In fact, the ability to tag friends in photos is one of the more popular features on

A. AWS

B. Facebook

C. Google

D. Tensorflow

B. Facebook

[QUES] An interesting application of image stitching is the ability to taken with a panning camera

A. multiple-image stitching

B. summarize and compress videos

C. panoramic video textures

D. Videobrush system

C. panoramic video textures

[QUES] One of the simplest (and most fun) applications of image alignment is a special form of image stitching called

A. panography

B. Whiteboard and document scanning

C. plane at infinity

D. RANSAC

A. panography

[QUES] 2. What methodology is presented to detect rust on iron bridges?

A. A person on the ground takes multiple high-resolution images from the same place. A computer vision expert splits these images into smaller groups. Each of the smaller images is passed to a custom classifier that can detect the presence of the metal structure.

B. A person on the ground takes multiple high-resolution images from multiple places. A computer vision expert splits these images into smaller groups. Each of the smaller images is passed to a custom classifier that can detect the presence of the metal structure.

C. A person on the ground takes multiple high-resolution images from the same place. A computer vision expert splits these images into smaller groups. Each of the smaller images is passed to a custom classifier that can detect the presence of the metal structure versus other, non-metal structures. After this, the images are passed through another custom classifier that is trained to detect the presence of rust in images.

D. A person on the ground takes multiple high-resolution images from multiple places. A computer vision expert splits these images into smaller groups. Each of the smaller imag

D. A person on the ground takes multiple high-resolution images from multiple places. A computer vision expert splits these images into smaller groups. Each of the smaller images is passed to a custom classifier that can detect the presence of the metal structure versus other, non-metal structures. After this, the images are passed through another custom classifier that is trained to detect the presence of rust in images.

[QUES] 3. The popularity of self-driving cars has been rising at an exponential rate over the past decade. Based upon what you have learned, which of the following computer vision technique(s) is useful for self-driving cars? Select all relevant answers

A. Object Detection

B. Motion Transfer

C. Image Classification

D. All of the above

C. Image Classification

[QUES] 4. Bob works in a pharmaceutical company where his job is to monitor the quality of the tablets produced. Broken or partially formed tablets need to be separated from the correctly produced ones before the tablets are carefully counted and packed in containers. Based upon what you've learned so far, what would you suggest Bob do?

A. Tell him to manually inspect each tablet produced and remove the deformed tablets from the production line.

B. Take pictures of the containers containing all the packed tablets right before the containers are ready to be shipped. Then use computer vision techniques on these pictures to separate the good tablets from the bad ones.

C. Hire you to manually handpick each tablet and separate the bad ones from the good ones.

D. Take pictures of all the tablets that are produced in the assembly line before they are sent to be packaged in containers. Use computer vision techniques on these pictures to separate the good tablets from the bad ones which prevent the bad tablets from getting shipped.

D. Take pictures of all the tablets that are produced in the assembly line before they are sent to be packaged in containers. Use computer vision techniques on these pictures to separate the good tablets from the bad ones which prevent the bad tablets from getting shipped.

[QUES] 5. Security at the airport is a topic of global concern. We don't want a miscreant carrying dangerous items onto airplanes that could cause potential harm to the lives of other passengers. How can we use Computer Vision to prevent people from carrying hazardous items onto airplanes?

A. Get more police dogs to detect hazardous materials.

B. Install scanners for scanning luggage and people. Use a custom classifier that is trained to identify hazardous items in images.

C. Hire more security guards at the airport security.

D. Install scanners for scanning luggage and people. Use a custom classifier that is trained on images to identify regular everyday items.

B. Install scanners for scanning luggage and people. Use a custom classifier that is trained to identify hazardous items in images.

[QUES] 6. What type of image operation can convolution perform?

A. Edge detection

B. Sharpening

C. Blurring

D. All of the above

D. All of the above

[QUES] 7. What is linear filtering?

A. It is a standard way to add text data

B. It is a standard way to filter Images using convolution

C. It is a standard way to filter text

D. None of the above

B. It is a standard way to filter Images using convolution

[QUES] 8. A video sequence is a:

A. Sequence of images

B. a large image

A. Sequence of images

[QUES] 9. OpenCV:

A. has more functionality than PIL library, but is more difficult to use

B. identical to PIL

A. has more functionality than PIL library, but is more difficult to use

[QUES] 10. In OpenCV an image is a:

A. a numpy array, with intensity values as 8-bit unsigned

B. an image object

A. a numpy array, with intensity values as 8-bit unsigned

[QUES] 11. An application of the ImageOps module is:

A. In OpenCV and flips images

B. In PIL, change colors

C. In PIL as flip and mirror functions

D. In OpenCV, adjust contrast and brightness

C. In PIL as flip and mirror functions

QUES] 12. Translation is:

A. Shrinking or expanding the image in a horizontal and/or vertical direction

B. Shifting the image

C. Rotating the images

D. None of the above

B. Shifting the image

[QUES] 13. The height of the image is the:

A. Number of columns

B. Number of rows

C. Number of pixels

D. None of the above

B. Number of rows

[QUES] 14. The order of channels in OpenCV:

A. BRG

B. GRB

C. RGB

D. BGR

A. BRG

[QUES] 15. The main points that we must remember when we work with OpenCV are:

A. 2

B. 4

C. 3

D. 9

C. 3

[QUES] 16. If we want to draw a shape and fill it inside we must write:

A. 1

B. fillnside = 1

C. fillnside = True

D. -1

D. -1

[QUES] 18. We have open an image with Matplotlib.pyplot and we want to remove only the Blue channel. Our code will be:

A. [:, :, 1] = 0

B. [:, :, 1] = -1

C. [:, :, 2] = 0

D. [:, :, 2] = -1

C. [:, :, 2] = 0

[QUES] 20. We are working with OpenCV and we want to draw a straight line on our image. How many points do we need to use?

A. 2

B. 4

C. 5

D. 3

A. 2

[QUES] 21. We have opened an image with OpenCV. We want to flip it on both Vertical and Horizontal axis. Which is correct the code ?

A. cv2.flip(img, -1)

B. cv2.flip(img, 0)

C. cv2.flip(img, 1)

D. cv2.flip(img, 2)

A. cv2.flip(img, -1)

[QUES] 23. You select the learning rate:

A. The learning rate should always be one

B. Using the validation data

C. Randomly

B. Using the validation data

QUES] 24. Support Vector Machines for Image classification may use kernels. There are different types of Kernels, which of the following is not a type of Kernel?

A. Histogram of Oriented Gradients

B. Radial Basis Function

C. Polynomial

D. Linear

A. Histogram of Oriented Gradients

[QUES] 25. The difference between Logistic Regression and Softmax is

A. Softmax is used for Multi-class Classification

B. Only Logistic Regression has probabilistic outputs

C. They are the same

A. Softmax is used for Multi-class Classification

[QUES] 26. Other methods to convert a two-class classifier to multiclass classifier they include

A. One-vs-rest

B. One-vs-one

C. SVM

B. One-vs-one

[QUES] 27. What methods can you use for Multiclass classification with no alteration

A. KNN

B. Logistic Regression

C. SVM

D. Softmax

D. Softmax

[QUES] 28. You train a Support Vector Machine with a RBF kernel and obtain an accuracy of 100% on the training data and 50% on the validation data. What should you do to the parameter Gamma?

A. Increase Gamma

B. Decrease Gamma

C. Leave Gamma unchanged

B. Decrease Gamma

[QUES] 29. When classifying images into tiger and lions, which image classification algorithm finds the best plane that divides a dataset into two classes?

A. KNN

B. Logistic Regression

C. SVM

D. K-Means

C. SVM

[QUES] 30. Which of the following generate Histograms for each region of an image separately using gradients and the orientation of pixel values?

A. SVM

B. KNN

C. Softmax

D. Histogram of Oriented Gradients

D. Histogram of Oriented Gradients

[QUES] 31. You train a Support Vector Machine and obtain an accuracy of 100% on the training data and 50% on the validation data. This is an example of:

A. Overfitting

B. Underfitting

C. A good model

A. Overfitting

[QUES] 32. When dealing with image classification, what kind of challenges do we face with images?

A. Variations due to scaling

B. Image occlusion

C. Variations due to illumination

D. All the above

D. All the above

[QUES] 33. How many colour channels does a Grayscale image have?

A. 1

B. 2

C. 3

D. 4

A. 1

[QUES] 34. Which of these are a problem of sliding windows? Select all that apply

A. Aspect Ratio

B. Object sizes

C. Overlapping objects

D. Grayscale image

C. Overlapping objects

[QUES] 35. P. Viola and M. Jones based the idea of the Cascade classifier on which of the following?

A. Deep learning

B. Image feature extraction

C. Haar Wavelet Sequence

D. Sliding window sequence

C. Haar Wavelet Sequence

[QUES] 36. In object detection the score:

A. Gives you the probability of each class

B. is bounding box region

C. Is the confidence of the model prediction

D. Saves it for a different algorithm

C. Is the confidence of the model prediction

[QUES] 37. When we are dealing with object detection, there are many different classifiers that we can use. Which of the following classifiers is trained on a large number of images that include the object we are trying to detect as well as images that do not contain the object we are trying to detect?

A. Cascade Classifiers

B. Sliding window Classifiers

C. Viola Classifiers

D. Integral Classifers

A. Cascade Classifiers

[QUES] 38. Which of the following is not accurate about the AdaBoost?

A. When using an Adaboost, it focuses on the correctly classified examples in the following round

B. In an Adaboost, a strong classifier is a linear combination of a weak classifier

C. Adaboost selects only those features that help to improve the classifier accuracy

D. Adaboost cuts down features of an image significantly

A. When using an Adaboost, it focuses on the correctly classified examples in the following round

[QUES] 40. Which of the following is the first fundamental step in image processing?

A. Image acquisition

B. Filtration

C. Image restoration

D. Image enhancement

A. Image acquisition

[QUES] 41. What is the translation transformation?

A. It changes the size of an object

B. It is scaling, skewing, and rotation

C. We rotate the object at a particular angle

D. It moves an object to a different position on the screen

D. It moves an object to a different position on the screen